Ticket #16079: 16079.3.2.diff

wp-includes/formatting.php

@@ -2066 +2066 @@
                 $more = __( '…' );
         $original_text = $text;
         $text = wp_strip_all_tags( $text );
-        $words_array = preg_split( "/[\n\r\t ]+/", $text, $num_words + 1, PREG_SPLIT_NO_EMPTY );
-        if ( count( $words_array ) > $num_words ) {
-                array_pop( $words_array );
-                $text = implode( ' ', $words_array );
-                $text = $text . $more;
-        } else {
-                $text = implode( ' ', $words_array );
+        
+        /* translators: How would you like to count Latin (English, French, etc.)?  
+                 translate this to 'characters' if you think 'two words' shoule be counted as 7,  
+                 translate this to 'words' if you think 'two words' should be counted as 2. 
+                 Default: 'words'. 
+        */
+        $count_latin_by = 'characters' == _x( 'Count Latin by', 'Latin word count: as words or characters?' ) ? 'characters' : 'words';
+        
+        /* translators: If you set countLatinBy to 'characters', would you count spaces? 
+                 translate this to 'yes' if you want spaces to be counted, 
+                 translate this to 'no' if you do not want spaces to be counted. 
+                 Default: 'no'.
+        */
+        $count_spaces = 'yes' == _x( 'Count Latin spaces', 'Latin spaces count: yes or no?' ) ? true : false;
+        
+        /* translators: Would you like to count punctuation marks for East Asia text? 
+                 translate this to 'yes' if you want punctuation marks to be counted, 
+                 translate this to 'no' if you do not want punctuation marks to be counted.
+                 Default: 'yes'.
+        */
+        $count_eastasia_punc = 'no' == _x( 'Count East Asia punctuation marks', 'East Asia punctuation marks count: yes or no?' ) ? false : true;
+        
+        /* translators: Would you like to cut a Latin word apart to fit the word limit? 
+                 translate this to 'yes' if you allow a word to be cut if it is too long, 
+                 translate this to 'no' if you do not allow a word to be cut if it is too long.
+                 Default: 'no'.
+        */
+        $break_words = 'yes' == _x( 'Break Latin words when trimming text to a certain number of words', 'For a long word, should I break it apart to fit the word limit: yes or no?' ) ? true : false;
+        
+        @mb_internal_encoding( get_option( 'blog_charset' ) );
+        
+        $subject = $text;
+        $text = '';
+        
+        $preg_eastasia_char = '[\x{3100}-\x{312F}\x{31A0}-\x{31BF}\x{4E00}-\x{9FCF}\x{3400}-\x{4DBF}\x{F900}-\x{FAFF}\x{2F00}-\x{2FDF}\x{2E80}-\x{2EFF}\x{31C0}-\x{31EF}\x{2FF0}-\x{2FFF}\x{1100}-\x{11FF}\x{A960}-\x{A97F}\x{D780}-\x{D7FF}\x{3130}-\x{318F}\x{FFA0}-\x{FFDC}\x{AC00}-\x{D7AF}\x{3040}-\x{309F}\x{30A0}-\x{30FF}\x{31F0}-\x{31FF}\x{FF65}-\x{FF9F}\x{3190}-\x{319F}\x{A4D0}-\x{A4FF}\x{A000}-\x{A48F}\x{A490}-\x{A4CF}]';
+        $preg_eastasia_punc = '[\x{3000}-\x{303F}\x{FE30}-\x{FE4F}\x{FF01}-\x{FF60}\x{FE10}-\x{FE1F}]';
+        $preg_latin = '[A-Za-z0-9\x{0080}-\x{00FF}\x{0100}-\x{017F}\x{0180}-\x{024F}\x{2C60}-\x{2C7F}\x{A720}-\x{A7FF}\x{1E00}-\x{1EFF}\x{0027}\x{2019}\x{2010}-\x{2015}-]';
+        $preg_latin_punc = '[0-9.(),;:!?%#$¿\'"_+=\\/-]';
+
+        $i = 0;
+        $words = 0;
+        while ( true ) {
+                // Should anything fails, this prevents us from going into an infinite loop
+                if ( $i++ >= $num_words * 10 )
+                        break;
+        
+                if ( trim( $subject ) == '' )
+                        break;
+        
+                if ( $words >= $num_words )
+                        break;
+        
+                $fragment_words = 0;
+        
+                $subject = preg_replace_callback(
+                        "/^(?:$preg_eastasia_char|$preg_eastasia_punc|(?<Latin>$preg_latin+$preg_latin_punc*)|.)\s*/u",
+                        create_function(
+                                '$matches',
+                                'global $wp_trim_words_fragment_matches; $wp_trim_words_fragment_matches = $matches; return \'\';'
+                        ),
+                        $subject,
+                        1
+                );
+        
+                // XXX workaround to get matches from preg_replace_callback()
+                global $wp_trim_words_fragment_matches;
+                $fragment = $orig_fragment = $wp_trim_words_fragment_matches[0];
+                $is_fragment_latin = isset( $wp_trim_words_fragment_matches['Latin'] ) ? true : false;
+        
+                if ( $is_fragment_latin && $count_latin_by == 'words' ) {
+                        $text .= $orig_fragment;
+                        $words++;
+                        continue;
+                }
+        
+                if ( ! $count_eastasia_punc )
+                        $fragment = preg_replace( "/$preg_eastasia_punc/u", '', $fragment );
+        
+                if ( ! $count_spaces )
+                        $fragment = trim( $fragment );
+        
+                $fragment_words += mb_strlen( $fragment );
+        
+                if ( $words + $fragment_words > $num_words ) {
+                        if ( ! $break_words || $count_latin_by == 'words' )
+                                break;
+                
+                        $fragment_chars = preg_split('/(?<!^)(?!$)/u', $orig_fragment );
+                
+                        $j = 0;
+                        $fragment_words = 0;
+                        while ( true ) {
+                                if ( ++$j > mb_strlen( $orig_fragment ) )
+                                        break 2;
+                        
+                                if ( $words + $fragment_words >= $num_words )
+                                        break 2;
+                        
+                                $text .= $fragment_chars[0];
+                                $fragment_words++;
+                                $words++;
+                                array_shift( $fragment_chars );
+                        }
+                }
+        
+                $text .= $orig_fragment;
+                $words += $fragment_words;
         }
+
+        // Prevent other scripts from accessing fragments of users' post
+        unset( $GLOBALS['wp_trim_words_fragment_matches'] );
+        
+        if ( $words > $num_words )
+                $text .= $more;
+        
         return apply_filters( 'wp_trim_words', $text, $num_words, $more, $original_text );
 }